The authors describe a general-purpose, representation-independent method for the accurate and computationally efficient registration of 3-D shapes including free-form curves and surfaces. The method handles the full six degrees of freedom and is based on the iterative closest point (ICP) algorithm, which requires only a procedure to find the closest point on a geometric entity to a given point. The ICP algorithm always converges monotonically to the nearest local minimum of a mean-square distance metric, and the rate of convergence is rapid during the first few iterations. Therefore, given an adequate set of initial rotations and translations for a particular class of objects with a certain level of 'shape complexity', one can globally minimize the mean-square distance metric over all six degrees of freedom by testing each initial registration. One important application of this method is to register sensed data from unfixtured rigid objects with an ideal geometric model, prior to shape inspection. Experimental results show the capabilities of the registration algorithm on point sets, curves, and surfaces. >

A method for registration of 3-D shapes

We describe an object detection system based on mixtures of multiscale deformable part models. Our system is able to represent highly variable object classes and achieves state-of-the-art results in the PASCAL object detection challenges. While deformable part models have become quite popular, their value had not been demonstrated on difficult benchmarks such as the PASCAL data sets. Our system relies on new methods for discriminative training with partially labeled data. We combine a margin-sensitive approach for data-mining hard negative examples with a formalism we call latent SVM. A latent SVM is a reformulation of MI--SVM in terms of latent variables. A latent SVM is semiconvex, and the training problem becomes convex once latent information is specified for the positive examples. This leads to an iterative training algorithm that alternates between fixing latent values for positive examples and optimizing the latent SVM objective function.

/pdf/object-detection-with-discriminatively-trained-part-based-4cnosvuqcp.pdf

Object Detection with Discriminatively Trained Part-Based Models

Stereo matching is one of the most active research areas in computer vision. While a large number of algorithms for stereo correspondence have been developed, relatively little work has been done on characterizing their performance. In this paper, we present a taxonomy of dense, two-frame stereo methods designed to assess the different components and design decisions made in individual stereo algorithms. Using this taxonomy, we compare existing stereo methods and present experiments evaluating the performance of many different variants. In order to establish a common software platform and a collection of data sets for easy evaluation, we have designed a stand-alone, flexible C++ implementation that enables the evaluation of individual components and that can be easily extended to include new algorithms. We have also produced several new multiframe stereo data sets with ground truth, and are making both the code and data sets available on the Web.

/pdf/a-taxonomy-and-evaluation-of-dense-two-frame-stereo-4c20etkubp.pdf

A taxonomy and evaluation of dense two-frame stereo correspondence algorithms

A comparative study of texture measures with classification based on featured distributions

Planning algorithms are impacting technical disciplines and industries around the world, including robotics, computer-aided design, manufacturing, computer graphics, aerospace applications, drug design, and protein folding. This coherent and comprehensive book unifies material from several sources, including robotics, control theory, artificial intelligence, and algorithms. The treatment is centered on robot motion planning but integrates material on planning in discrete spaces. A major part of the book is devoted to planning under uncertainty, including decision theory, Markov decision processes, and information spaces, which are the “configuration spaces” of all sensor-based planning problems. The last part of the book delves into planning under differential constraints that arise when automating the motions of virtually any mechanical system. Developed from courses taught by the author, the book is intended for students, engineers, and researchers in robotics, artificial intelligence, and control theory as well as computer graphics, algorithms, and computational biology.

Planning Algorithms: Introductory Material

This dissertation concerns the mathematical theory of screw-transform manifolds and their use in camera self calibration. A camera's calibration is the function that maps 3D scene points to 2D image points, e.g., in photographs taken by the camera. Between every two photographs taken from different positions there exists a pairwise constraint called the “fundamental matrix,” which can be computed directly from the images. When the two photographs are captured by the same camera, the fundamental matrix induces a surface in calibration space called a “screw-transform manifold.” This manifold represents every possible internal calibration for the camera. By acquiring several different pairwise fundamental matrices, several different screw-transform manifolds can be computed; however, the internal calibration of the original camera must be a member of each manifold and hence, by finding the intersection point of all the manifolds, the camera's calibration can be determined. The process of determining calibration directly from images taken by a camera is called “self calibration.” 
The contributions of this dissertation include the theory of screw-transform manifolds and three original algorithms for determining the mutual intersection points of a collection of manifolds. While many papers have been written on self calibration, almost all previous methods posed their solutions as the global minima of an error function. However, performing global optimization is problematic; it is easy to locate a local minimum without finding the global minimum, and in some cases the attraction basin of the global minimum is so small that the algorithm must essentially “guess” the solution in order to find it. One of the new approaches created as part of this dissertation, called STM-SURFIT, avoids global optimization altogether and can effectively locate all global minima in a single pass, running in under 1 second on a modern home computer. The general approach used to avoid the problems of optimization may have wider applicability than simply camera calibration. 
A tutorial on multiview geometry that assumes only knowledge of linear algebra is included to provide the necessary mathematical background. The related history and previous work on self calibration and image-based rendering is also presented. As part of the theory of screw-transform manifolds, a theorem is introduced that partitions monocular view pairs into six categories based on the underlying screw motion of the camera and provides a simple test for determining category. In addition, some methods for self calibration and image-based rendering from dynamic scenes are presented. The image-based rendering techniques do not require camera calibration but are limited in applicability; this adds to the growing body of evidence that camera calibration is a necessity for most useful image-based rendering techniques.

Screw-transform manifolds for camera self calibration

Relation of one-way parallel/sequential automata to 2-D finite-state automata

In this paper we present the geometry and the algorithms for organizing and using a viewer-centered representation of the occluding contour of polyhedra. The representation is computed from a polyhedral model under orthographic projection for all viewing directions. Using this representation, we derive constraints on viewpoint correspondences between image features and model contours. Our results show that the occluding contour, computed in a model-based framework, can be used to strongly constrain the viewpoints where a 3D model matches the occluding contour features of the image.

An occlusion-based representation of shape for viewpoint recovery

A new representation of faces, called face cyclographs, is introduced for face recognition that incorporates all views of a rotating face into a single image. The main motivation for this representation comes from recent psychophysical studies that show that humans use continuous image sequences in object recognition. Face cyclographs are created by slicing spatiotemporal face volumes that are constructed automatically based on real-time face detection. This representation is a compact, multiperspective, spatiotemporal description. To use face cyclographs for recognition, a dynamic programming based algorithm is developed. The motion trajectory image of the eye slice is used to analyze the approximate single-axis motion and normalize the face cyclographs. Using normalized face cyclographs can speed up the matching process. Experimental results on more than 100 face videos show that this representation efficiently encodes the continuous views of faces.

/pdf/face-cyclographs-for-recognition-45qvz1dqq9.pdf

Face Cyclographs for Recognition

A nonparametric Bayesian model for attribute-based object recognition and image-based class attribute inference is presented. This model draws on existing work in Bayesian nonparametrics such as the focused topic model [25]. The model allows either the classification of objects or the inference of attributes over known classes (or both simultaneously). Attributes inferred from image datasets allow an improvement in classification accuracy when combined with attributes from other sources, including “layperson” knowledge and existing inference methods.

Charles R. Dyer

Papers

Screw-transform manifolds for camera self calibration

Relation of one-way parallel/sequential automata to 2-D finite-state automata

An occlusion-based representation of shape for viewpoint recovery

Face Cyclographs for Recognition

The Multimodal Focused Topic Model: A Nonparametric Bayesian Approach to Simultaneous Object Classification and Attribute Discovery